Delivery of gentamicin by a controlled-release infusion system versus a minibag system

Delivery of gentamicin via a new controlled-release intravenous infusion system was compared with conventional delivery via small-volume injections in minibags by measuring serum drug concentrations in 10 healthy men. Each volunteer received gentamicin (as the sulfate salt) 2 mg/kg. In phase 1, subjects randomly received the drug either as a 50-mL admixture in 5% dextrose injection (D5W) or from the controlled-release system (CRIS, IVAC Corporation), in which drug was diluted in a vial with 10 mL of sterile water for injection (density of drug solution, approximately 1.5% w/v) and was delivered when the primary solution (D5W; density, 5% w/v) displaced drug from the vial and infused it into the subject over 30 min; subjects were then crossed over. In phase 2, nine of the subjects received the drug via CRIS with the diluent changed to 10 mL of 5% dextrose and 0.9% sodium chloride injection (D5NS; density of drug solution, approximately 5.9% w/v). In phase 3, 10 men (seven of the original subjects) received the gentamicin dose via CRIS with 20 mL of D5NS as the diluent or via minibags in a crossover design. The amount of drug remaining in each vial used with the CRIS system was determined. Drug administration via CRIS with 10 mL of sterile water diluent resulted in serum concentrations approximately 35% of those obtained with the minibag system, and a substantial portion (71 ± 8%) of the dose to be administered remained in the vials. With D5NS 10 mL as the diluent, less than 1% of the dose remained in the vials and postdistribution serum gentamicin concentrations were comparable to those for minibag doses. Approximately 4% of the dose prepared in 20 mL of D5NS remained in the CRIS vials. If the manufacturer's recommendations for use are followed, CRIS may be satisfactory for delivery of small doses of potent drugs. The diluent should be chosen so that the drug solution will be more dense than the primary solution being infused.